This content has been downloaded from IOPscience. Please scroll down to see the full text.

Download details:

IP Address: 89.187.219.181

This content was downloaded on 06/11/2015 at 17:43

Please note that terms and conditions apply.

RBS and XRD Characterization of Yttrium Iron Garnet Thin Films

View the table of contents for this issue, or go to the journal homepage for more

2015 J. Phys.: Conf. Ser. 592 012138

(http://iopscience.iop.org/1742-6596/592/1/012138)

Home Search Collections Journals About Contact us My IOPscience

RBS and XRD Characterization of Yttrium Iron Garnet Thin

Films

M MANSOUR1, M ROUMIE

2, B ABDEL SAMAD

3, H BASMA

1, M KOREK

1

1Faculty of Science, Beirut Arab University, Beirut, Lebanon

2 National Council for Scientific Research (CNRS), Beirut, Lebanon

3Laboratoire DIOM, Université de Lyon, Saint-Étienne, France

Email: mansourmahdy@live.com.

Abstract. Magnetic materials such as yttrium iron garnet (YIG or Y3Fe5O12) present a great

importance for their magneto-optic properties. They are potential materials used for

applications in the domain of optical telecommunications for example. In this work, we have

investigated YIG thin films deposited on substrates of quartz and GGG (gadolinium gallium

garnet or Gd3Ga5O12). Using Rutherford backscattering spectrometry (RBS) we characterized

the performed layers (thickness and stoichiometry) in order to correlate the films preparation conditions with the quality of the final material. We determined the optimal energy of the alpha

particles beam used for RBS measurements and we fitted the experimental spectra using the

SIMNRA simulated code. Our RBS results showed that the films have a stoichiometry close to

that of the starting material. In addition, we found that the film thickness is proportional to

deposition time but inversely proportional to the substrate temperature. Moreover, using x-ray

diffraction (XRD) we determined the annealing effect on the structure of the profile of our thin

films.

1. Introduction Recent studies conducted on the preparation of YIG thin films have shown its importance especially in

fields related to the field of magnetooptical applications in microwave frequency domains [1, 2].The

YIG is a ferrimagnetic garnet crystal with the composition of Y3Fe5O12, with a substantial Faraday rotation in large parts of the optical and microwave spectrum [3]. In our work, YIG thin films were

grown using the technique of radio frequency (RF) magnetron sputtering [2-4].This technique possess

several advantages as being a dry process during deposition, and having possibility of high purity

starting material, ability to sputter dielectric materials, commonly held industrial process and high compatibility with semiconductors technology. The radio frequency sputtering films were grown on

quartz and GGG substrates. The Rutherford back scattering technique (RBS) has proven to be very

effective for characterization of the thickness and stoichiometry of thin films [5, 6] having an advantage of being nondestructive. This technique allows us to correlate the quality of the films to the

different growth conditions and choose the optimal conditions. The x ray diffraction technique (XRD)

serves as a powerful tool to investigate any variation in the structure of the prepared films due to the

annealing process [7].

2. Experimental Techniques: The samples were prepared by radio frequency (RF) magnetron sputtering. Table.1 shows the different

sputtering conditions for different prepared samples.

International Conference on Strongly Correlated Electron Systems 2014 (SCES2014) IOP PublishingJournal of Physics: Conference Series 592 (2015) 012138 doi:10.1088/1742-6596/592/1/012138

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distributionof this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.

Published under licence by IOP Publishing Ltd 1

Table.1 Sputtering conditions for different prepared samples

Target Y3Fe5O12

Substrate Quartz or GGG

Argon flow 50sccm

Input power 100W

Sputtering time between1h30 and 3h00

Annealing temperature 740 °C

Annealing atmosphere Air

Annealing time 2h

Substrate position 4.6cm

The Rutherford back scattering measurements were performed using a 5SDH pelletron tandem

accelerator of 1.7 MV located at the Lebanese Atomic Energy Commission under normal incident

beam and in a random direction to avoid channeling. The measurements were done by using two

different energies of alpha-particles beam: 2 MeV and 3.5 MeV. In our experiments, X-ray measurements were done using a D8 Discover X-ray diffractometer from

Bruker AXS systems at the Central Research Laboratory (CRSL) in the American University of Beirut

(AUB). The X-ray tube emits radiation of wavelength 5418.1 Å.

3. Results and discussions:

The different RBS spectra were processed with the SIMNRA simulation code [8]. Indeed, RBS measurements are insensitive to the chemical state of the atoms and the quantitative interpretation of

the measurement results is absolute and in most cases unambiguous, as the physical processes

underlying ion-matter interactions are well understood [9]. The use of the classical RBS at 2 MeV He

++ beam was not useful for some of the samples that have thicker YIG layer. However, the use of

alpha particle beam at 3.5 MeV, with 5 µC fluence, was enough to determine the YIG layer thickness

for all prepared samples. Usually, the sensitivity on the thickness determination for a thin film is better at 2 MeV than 3.5 MeV, as the electronic stopping power (dE/dx)e is higher. However, when

comparing the thickness of some of the thinner YIG layers obtained by both energies, the difference in

values was found to be less than 3%. Besides, especially for GGG substrate, the O yield in the 2 MeV

RBS spectra was overlapped by those of Gd and Ga. Hence, the use of 3.5 MeV enhanced the oxygen yield, since at this energy the oxygen cross section is non-Rutherford, while those of Ga and Gd are

still Rutherford. The main results obtained from the RBS measurements after comparing the two

substrates are that, the two substrates, the film thickness showed a similar trend and almost the same values were obtained for samples taken in the same conditions (Table 2). One can see that the

thickness is proportional to the deposition time when comparing yq1 vs. yq2 and yg1 vs. yg2 grown

during 180 minutes and 90 minutes, respectively. However, the general tendency is to be confirmed in

future work by considering other values of the deposition times.

International Conference on Strongly Correlated Electron Systems 2014 (SCES2014) IOP PublishingJournal of Physics: Conference Series 592 (2015) 012138 doi:10.1088/1742-6596/592/1/012138

2

Fig. 1 illustrates the variation of the film thickness versus substrate temperature for quartz and GGG.

There is a decreasing trend with increasing temperature so the thickest layers were obtained at room

temperature.

For quartz substrate, Fig.2 shows the RBS spectrum of the yq2 samples and the variation of the

stoichiometric ratio of cation (Y+Fe)/anion (O) of the YIG films with the substrate temperature. The

chemical composition of the films prepared at T= 25°C, 500°C and 800°C are Y2.51Fe5.42O12, Y3.51Fe5.35O12 and Y3.66Fe5.85O12, respectively. At 500°C and 800°C, the sputter yield of the positive

ions (Y+Fe) is higher than that of the negative ions (O), so the ratios cation/anion of the films (0.74

and 0.79 respectively) are higher than that of the target (0.67). In a general manner, the room

Table.2 Thickness, determined by RBS, of the YIG film deposited on quartz (yq)

and GGG (yg) as function of the deposition time and the substrate temperature

Sample

No

Substrate

temperature (°C)

Deposition

time (min.)

Film

thickness (μm)

yq0 25 90 890

yq1 500 180 1137

yq2 500 90 830

yq3 800 90 727

yg0 25 90 900

yg1 500 180 1180

yg2 500 90 812

yg3 800 90 785

Figure 1. Variation of the thickness of the

deposited YIG film on quartz and GGG substrates

as a function of the substrate temperature.

Figure.1 Variation of the thickness of the

deposited YIG film on quartz and GGG substrates

as function of the substrate temperature.

Figure.1 Variation of the thickness of the

deposited YIG film on quartz and GGG substrates

as function of the substrate temperature.

700

750

800

850

900

950

0 200 400 600 800 1000

Temperature (°C)

Th

ick

ness

(n

m)

Quartz substrate

GGG substrate

International Conference on Strongly Correlated Electron Systems 2014 (SCES2014) IOP PublishingJournal of Physics: Conference Series 592 (2015) 012138 doi:10.1088/1742-6596/592/1/012138

3

temperature sample is the closest to the stoichiometric YIG target and the ratios (Y+Fe)/O, Y/O and

Fe/O are increasing when the substrate temperature increases.

Fig.3 shows the RBS spectrum of the yg0 sample prepared at room temperature and the variation of the

different cation/anion ratio. A similar behaviour of the quartz substrate is also observed for the GGG

substrate. The chemical composition of the films prepared at T= 25°C, 500°C and 800°C are

Y3.07Fe5.24O12, Y3.38Fe5.32O12 and Y3.50Fe5.47O12, respectively. At room temperature, the chemical composition of the YIG film is the nearest to the stoichiometric YIG target. In general, the YIG films

elaborated on GGG substrate are closer to the stoichiometric YIG target than those elaborated on

quartz substrate.

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0 200 400 600 800 1000

Temperature (°C)

Sto

ich

iom

etr

ic r

ati

o (Fe+Y)/O

Fe/O

Y/O

Figure 2. Shows the experimental and simulated RBS spectra obtained for the

film grown on quartz substrate sample yq2 showing the edge of Y, Fe, Si and O and

the variation of the chemical stoichiometry (expressed as cation/anion ratio) of the

film with quartz substrate temperature.

Figure 3. Shows the experimental and simulated RBS spectra obtained for the film grown on GGG substrate sample yg0 showing the edge of Y, Fe, Gd, Ga and O and the

variation of the chemical stoichiometry (expressed as cation/anion ratio) of the film

with GGG substrate temperature.

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 200 400 600 800 1000

Temperature (°C)

Sto

ich

iom

etri

c ra

tio (Fe+Y)/O

Fe/O

Y/O

International Conference on Strongly Correlated Electron Systems 2014 (SCES2014) IOP PublishingJournal of Physics: Conference Series 592 (2015) 012138 doi:10.1088/1742-6596/592/1/012138

4

The GIXRD obtained spectra of YIG samples grown on quartz are shown in figure 4.The XRD results

show the dependence of the crystallinity structure with the substrate temperature.

4. Conclusiuon The application of the RBS technique was very helpful for the study of stoichiometry and

thickness of YIG thin films deposited, by RF sputtering, onto quartz and GGG substrates. This

study was necessary to clarify some trends of YIG film growth depending on the sample

preparation conditions. Indeed, it was found that the film thickness is proportional to the

deposition time while it is inversely proportional to the substrate temperature. The films

grown on GGG substrate are closer to the stoichiometric YIG than those grown on quartz and

a better stoichiometry is obtained at room temperature for both substrates. The preliminary

results of XRD show the dependance of the cristallinity structure with the substrate

temperature.

References: [1] Furuya A, Yoshikawa H,Tanabe T, Yamemoto M, Tailhades P,Bouet L, Despax C,

Presmanes L and Rousset A 1999, J. App. Phys, 85 5106

[2] Inoue M and Fujii T 1997, J. App. Phys. 81 8 [3] Boudiar T, Payet-Gervy B, Blanc-Mignon F, Rousseau J, Le Berre M,Joisten H 2004, J. Magn

284 77

[4] Furuya A, Yoshikawa H, Tanabe T, Yamemoto M, Tailhades P, Bouet L, Despax C,

Presmanes L and Rousset A 1999, J. Appl. Phys. 85 5106 [5] Boudiar T, Capraro S, Rouiller T, Blanc-Mignon M-F, Payet. Gervy B, Le Berre M, and

Rousseau J-J 2004, J Phys.Stat. Sol. C, 1, 3347

[6] Abdel Samad B et. Al 2010 Eur. Phys. J. Appl. Phys. 50 10502 [7] Park M-B, Cho N-H 2004, J.M.M.M , 231

[8] Mayer M 1997, Max-Planck-Institut fürPlasmaphysik, Garching, SIMNRA User's Guide

Report IPP 9/113 [9] Amsel G and Battistig G 2005, Nucl. Inst. and Meth. B 240

Figure 4. Shows the annealing effect on the structure of the profile of our samples. This

effect can be shown by comparing the results with and without annealing and the deposition temperature effect with annealing at 740

0C.

International Conference on Strongly Correlated Electron Systems 2014 (SCES2014) IOP PublishingJournal of Physics: Conference Series 592 (2015) 012138 doi:10.1088/1742-6596/592/1/012138

5